The NIMH Research Domain Criteria (RDoC) framework is a heuristic approach to integrate neuroscience with psychopathology. Thus, RDoC is the attempt to develop constructs that (a) connect across units of analysis, (b) are applicable across established diagnostic criteria, and (c) are a point of departure to provide measurable constructs that will be useful to gain a deeper understanding of psychiatric conditions. However, the lack of reliable, robust latent constructs derived from experimental measures is a critical barrier to organizing mental disorders along dimensions that can be mapped onto multiple units of analysis. We focus on the positive and negative valence system domains proposed by the RDoC using self-report, behavior, physiology, and neural circuit unit of analysis measures, apply them to a clinical population of individuals with anxiety or depression recruited from primary care clinics across two sites, and apply latent variable analysis to derive latent constructs of positiv and negative valence systems functioning that cut across units of analyses. The specific aims of the proposal are: (1) To recruit, assess, and analyze self-report, behavioral, physiological, and neural circuit unit of analysis data in an exploratory sample from a cohort of n=100 treatment seeking individuals presenting to primary care clinics at UCSD and UCLA. (2) To obtain a confirmatory sample of n=100 subjects at both sites using the same measurement approaches to determine the reproducibility and robustness of the latent constructs obtained from the exploratory sample. (3) To obtain a reliability sample of n=50 individuals at UCSD and UCLA that will be re-tested within a month to determine the reliability of the latent constructs. This wll be the first time neural indices of positive valence and negative valence are investigated within the same broadly defined sample of anxious or depressed patients. Extant studies of neural processes in emotional disorders are limited to relatively small sample sizes. By combining resources across two sites, we will recruit a large enough sample to conduct more comprehensive analyses than ever before achieved. Once the proposed aims are completed, we will have a robust and reliable dimensional set of variables that quantify the positive and negative valence domains based on a latent variable approach. These variables capture clinical variance and will have been obtained from two different clinical sites to provide empirical evidence of the generality of the negative and positive valence domains. These variables will subsequently be used to determine whether (a) they can be used as predictors for how well an individual will respond to pharmacological or behavioral treatment; (b) they can be differentially responsive to therapies that are specifically aimed at modifying the positive or negative valence systems, and (c) they can be used in subsequent computational and/or neural process models of anxiety and depression to gain a more fundamental understanding of the pathology that occurs in these individuals.